The field of the disclosure relates generally to context aware information collection and aggregation in a dynamically changing and distributed environment, where information is needed by different information consumers for different purposes. Specifically, the disclosure is directed to methods and systems for context-aware entity correspondence and merging.
Existing information aggregations are static and current approaches don't address dynamic information aggregation. In contrast, context is any information used in realization of the meaning of an entity and any information used to characterize the situation of an entity. Examples of context information may include: location, identity, time and activity. Context answers fundamental questions such as what is occurring in the situation, where you are, who you are with, and what objects are around. Additionally, context refers to the current values of specific ingredients that represent a user's activity/situation.
Entity resolution (ER) approaches do not analyze the context features. Specifically, a disambiguation quality associated with ER approaches frequently depends on the context in which they are employed. Another application of the proposed approach is that if the context can be captured then this knowledge can be utilized to significantly improve performance of an ER ensemble.